Generative Knowledge Search Engine for Enterprise Network Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional search engines in enterprise networks fail to capture latent correlations and provide solution-based outputs, struggling with complex network management due to fragmented knowledge and shallow customization, leading to overwhelming search results and difficulty in understanding user intentions.
Innovation Solution
A Generative Knowledge Search Engine (GSE) utilizing Graph Generative Pre-Trained Transformers (G-GPT) that contextualizes search queries within the enterprise's asset data and network portfolio, generating cohesive solution graphs that reflect user intentions and latent correlations, reducing computational costs and overwhelming results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional search engines are used in enterprise networks, then basic keyword search functionality is provided, but latent correlations between enterprise assets and services cannot be captured, leading to fragmented knowledge and shallow customization
Solution Approach 1:
The patent merges multiple enterprise assets (inventory data, network portfolio, service functions) into a unified knowledge graph that captures latent correlations between them. This integration allows the system to understand relationships across different enterprise components that traditional search engines cannot detect, resolving the contradiction by combining fragmented information sources into a cohesive structure.
Solution Approach 2:
The system performs preliminary contextualization of search queries against enterprise asset data before generating search results. By pre-processing queries with Graph Generative Pre-Trained Transformers to understand user intentions and contextualize them within the enterprise network, the system captures latent correlations upfront, avoiding the need for complex post-processing while maintaining information completeness.
2Ease of operation
If traditional search engines provide search results, then query responses are generated, but results are overwhelming and lack solution-based organization, making it difficult to understand user intentions
Solution Approach 1:
The patent extracts only the most relevant and solution-oriented information from the enterprise knowledge graph, filtering out unnecessary details. By using graph neural networks to identify and extract key relationships and solutions directly related to user queries, the system reduces the volume of search results while maintaining ease of operation, presenting only the most valuable information in an organized manner.
Solution Approach 2:
The system applies local quality by providing customized, context-specific search results tailored to each user's enterprise network environment. Rather than providing generic results, the search engine adapts the quality and relevance of results to the specific enterprise assets and services involved, making the information more actionable and easier to operate with for each specific case.
3Productivity
If enterprise networks are managed with traditional IT specialist teams, then basic network management is performed, but understanding enterprise network features and assets is difficult, and addressing IT issues is complicated
Solution Approach 1:
The patent introduces an AI-powered search engine as an intermediary between IT specialists and enterprise network assets. This intermediary translates complex network configurations and asset relationships into understandable insights, bridging the gap between human expertise and system complexity. The search engine acts as a mediator that simplifies the interaction between operators and the complex enterprise network, improving productivity without reducing the inherent complexity of network management.
Data Source
AI summary
Methods are provided for generating end-to-end solutions-based search results for multi-query search inquiry. The search results are generated using graph generative pre-trained transformers and a network knowledge base. A method involves obtaining at least one search query and inventory data that includes information about a plurality of enterprise assets and configuration of an enterprise network. The method further includes generating a contextual schema based on the inventory data. The contextual schema includes a plurality of query sub-graphs indicative of an intention of the at least one search query and generating a solution graph by performing machine learning with respect to the plurality of query sub-graphs and network domain knowledge data. The method further includes providing a response to the at least one search query based on the solution graph. The response is specific to the enterprise network.